透過您的圖書館登入
IP:3.238.12.0
  • 學位論文

考量極紫外光閃焰及化學機械研磨之電路擺置

Simultaneous EUV Flare- and CMP-Aware Placement

指導教授 : 張耀文

摘要


極紫外光微影是最有前途的下一代微影技術之一,但此微影技術遭受閃焰(散射光)效果的影響,導致關鍵尺寸的失真和均勻性的損失。此外,化學機械研磨消除氧化矽、金屬和多晶矽的表面差異,在現今的積體電路製造過程中是一個重要的過程。這些製造性的問題與佈局密度分布有著高度相關,然而,這兩種技術間最佳化的佈局密度分布存在著差異性,針對紫外光閃焰,佈局密度分布趨勢為晶片中央區域密度較高的凸型曲面可獲得較佳的閃焰最佳化結果,而針對化學機械研磨,保持密度分布的均勻性可獲得較佳的研磨成果。為了要控制佈局密度分布,電路擺置是一個關鍵的階段。電路擺置的結果將顯著影響金屬密度分布。在這篇論文中,我們提出了一個電路擺置的演算法可以同時考慮極紫外光閃焰、化學機械研磨以及線長。其中,我們提出了一個s型(sigmoid)的模型去逼近最佳的密度分布使得極紫外光閃焰的影響能得到減少,並且在電路擺置階段,為了評估金屬的密度,我們提出電路接腳模型來估算金屬密度分布,最後使用非線性的最佳化方法獲得良好的擺置成果。實驗結果顯示,我們提出的電路擺置演算法可以有效地和高效地同時最佳化極紫外光的閃焰影響、化學機械研磨和線長。

並列摘要


Extreme Ultraviolet Lithography (EUVL) is one of the most promising lithography technologies for next generation, but EUVL suffers from the flare effects (i.e., scattered light) which cause critical dimension (CD) distortion and damage CD uniformity. Moreover, Chemical Mechanical Polishing (CMP) is an important process which planarizes silicon oxide, metal, and polysilicon surfaces in modern IC manufacturing. Both of the manufacturability issues are highly related to layout pattern distribution. However, there is a trade-off between the two techniques because the desirable pattern distributions for EUV flare and CMP optimization are different. To minimize EUV flare effects, a concave density distribution is preferred, where the center of a chip has higher density than periphery regions. On the other hand, CMP optimization tends to keep the density distribution uniform. In addition to control layout pattern distribution, placement is a critical stage in VLSI design flow because the positions of circuit blocks would significantly affect metal distribution. In this thesis, we propose the first work of EUV flare- and CMP-aware placement that considers the two manufacturability issues and optimize wirelength simultaneously. Besides, we propose a sigmoid distribution model to approximate the optimal EUV flare-aware density distribution for EUV flare effect optimization. To evaluate the metal density during the placement stage, we present a metal-aware pin model to estimate metal distribution. At last, we use a non-linear optimization method to achieve good placement results. The experimental results show that our proposed algorithms can effectively and efficiently achieve the placement solutions with optimized EUV flare effects, the CMP cost, and wirelength.

參考文獻


[21] M.-K. Hsu, Y.-F. Chen, C.-C. Huang, and Y.-W. Chang, Routability-
[30] J. Lou, S. Thakura, S. Krishnamoorthy, and H. S. Sheng, Estimating routing
[28] K. S.-M. Li, Y.-W. Chang, C.-L. Lee, C. Su, and J. E. Chen, Multilevel full-
[10] T.-C. Chen, Z.-W. Jiang, T.-C. Hsu, H.-C. Chen, and Y.-W. Chang, NTU-
[3] A. R. Agnihotri, S. Ono, and P. H. Madden, Recursive bisection placement:

延伸閱讀